SpaceX Files for Direct-to-Cell License to Expand US Satellite Coverage

SpaceX filed its direct-to-cell license application to expand its US satellite connectivity operations. The direct-to-cell license application threatens to disrupt all national mobile network systems. The company uses its Starlink network, which consists of low-Earth-orbit satellites that provide internet access, to develop its operational capabilities.  

SpaceX intends for standard smartphones to link directly to satellites, without additional hardware. People who live in remote areas that currently lack mobile service will benefit from this technology because it enables them to access mobile networks. Satellite systems are being integrated into standard mobile networks according to this industry trend, which the company uses as its operational approach. 

Expanding Beyond Traditional Satellite Internet  

Starlink provides satellite internet as its primary service, delivering broadband internet to people living in remote areas. SpaceX has reached a critical development phase with the introduction of direct-to-cell wireless technology.  

The system enables mobile phones to establish satellite connections without needing cellular network infrastructure. The system allows users to send messages and use basic data functions even in areas without cellular towers. The industry is developing hybrid networks that combine satellite systems with ground-based networks to expand service coverage. 

How Direct-to-Cell Technology Works  

Direct-to-cell technology enables satellites to function as cell towers from space. They interact directly with mobile devices using dedicated radio frequencies (specific segments of the radio spectrum that phones use for communication) and advanced networking protocols (techniques for orchestrating and transmitting data).  

SpaceX plans to use spectrum bands (radio frequency segments already assigned for communications) compatible with common smartphones. No new equipment will be needed, making the system easy to use. The technology will first offer basic services like text messaging and emergency communication. SpaceX will later expand these offerings to include voice and data services.  

Addressing Coverage Gaps in the United States  

The case aims to resolve existing coverage gaps that impact all regions of the United States. The current cellular network system does not provide reliable service across extensive areas. The existing coverage system does not meet the needs of rural and remote areas. 

Direct-to-cell satellite services could bridge these gaps without costly infrastructure. The emergency situation requires greater coverage because responders need reliable communication systems to protect themselves. Satellite connections will enhance access where service is minimal.  

Integration with Existing Mobile Networks  

SpaceX partners with telecom networks instead of replacing them. Users access mobile services via direct-to-cell signals, even outside terrestrial coverage.  

Mobile carriers and regulatory authorities must work together to create rules governing spectrum allocation and equipment interoperability. Research shows that telecom provider partnerships will be essential for achieving broad usage of this technology.  

Engineers must establish network integration processes enabling legacy systems to interoperate with new devices while ensuring reliability. The communications network combines satellite and cellular technology for greater dependability and flexibility.  

Regulatory and Licensing Considerations  

The process of obtaining a direct-to-cell license for market entry establishes essential requirements. The system must prove its regulatory compliance while it protects its ongoing business operations.  

US government agencies that oversee spectrum management and telecommunications regulations conduct a comprehensive approval process that evaluates new technology against essential requirements for reliable communication systems.  

SpaceX submitted its application to demonstrate its intention to increase service availability, but the actual service start date will depend on regulatory decisions.  

Competitive Landscape and Industry Impact  

SpaceX is currently developing direct-to-cell connectivity technology, while other companies pursue similar research. Competition among companies will drive rapid technological advancements.  

Satellite providers will enter mobile connectivity markets, leading to operational changes and new industry partnerships across the entire telecommunications sector. The telecommunications sector will undergo major changes as a result of this process.  

Telecom operators will gain better coverage through satellite systems but must tackle three specific operational challenges: network integration, spectrum management, and changes to revenue streams. Businesses must adapt their strategies to market conditions to remain competitive while creating value for their stakeholders. 

Implications for Consumers and Businesses  

The direct-to-cell system, which launched in 2023, allows customers to access better communication services during phone calls in regions that experience poor network connections. The system operates as an emergency communication system. 

The agricultural sector, logistics sector, and energy sector will benefit from better connectivity, which will enhance their operational capabilities. Organizations require strong communication systems for their daily operations. The satellite systems deliver vital support, which enables their fundamental operations to function correctly. 

The technology enables new applications, such as Internet of Things (IoT) deployments and remote monitoring systems that require continuous network connectivity. IoT refers to networks of physical devices, such as sensors or machines, that communicate and share data. 

Challenges and Technical Limitations  

Direct-to-cell technology offers benefits to users, but it must address three challenges: latency, signal degradation, and limited satellite communication capacity. Developers need to create comprehensive testing procedures to assess system performance under different environmental conditions and operational scenarios. The system needs to prove its ability to handle increased operational requirements without losing performance quality. Engineers need to develop methods that will enable current systems to work together with new technology while preserving network security. 

A New Phase for Satellite Connectivity  

SpaceX’s direct-to-cell license application marks a major advance in satellite communication. The company is developing ways for satellites to connect directly with mobile devices. The technology has strong potential but depends on regulatory approval and resolution of technical issues. It could change how communications reach areas with unreliable service. Upcoming project developments will influence how companies build connectivity services in the US. These developments will also affect other countries. The outcome could redefine expectations for universal mobile coverage in the coming years.

Source: xAI joins SpaceX to Accelerate Humanity’s Future 

US Energy Department Deploys AI System to Automate Power Grid Operations

The American electrical grid is at a turning point. More renewable energy sources make manual supervision less effective. To manage this complexity, the U.S. Energy Department is now using an AI system to automate grid operations. This marks a shift away from reactive methods. The move is part of the Grid Modernization Initiative, aiming for a self-healing, active grid that makes decisions in milliseconds.  

For power systems, engineers, and policy stakeholders, this is more than a software update. It is a major change in grid stability. By 2026, the rapid growth of distributed energy resources such as home solar, EV charging, and battery storage creates multidirectional electricity flows, making traditional centralized control less effective.  

The Architecture Of Self-Driving Load Balancing 

At the center of the new Department of Energy (DOE) system is a detailed digital model of the North American power grid. The AI uses a decentralized network of edge-computing core nodes, small local computers that process information close to its source, to process data from millions of phasor measurement units (PMUs), devices that monitor electrical flow, and smart meters in real time. Unlike older supervisory control and data acquisition (SCADA) systems, which can have 5-minute delays, the new framework uses neural state prediction to anticipate electrical changes and predict voltage changes before they cause problems.  

The automation system uses a constrained reinforcement learning model. This lets the AI balance frequency regulation and generator ramping while staying within the transmission limits. For example, if clouds reduce solar power in the south-west or a cold front increases heating demand in the north-east, the AI acts autonomously. It shifts flexible nodes and activates virtual power plants to keep the frequency steady at 60Hz.  

Enhancing Protection From Cyber and Physical Threats 

One main reason for this deployment is the growing number of threats. The American power grid is a major target for both cyberattacks and severe weather driven by climate change. The DOE’s AI system uses a specialized anomaly-detection engine (software) that identifies unusual patterns to distinguish between real equipment failures and digital attacks.  

The AI monitors electromagnetic signals from substation equipment. It looks for signs of compromised circuit breakers or unauthorized relay commands. If an attack is confirmed, the system isolates the affected grid section. Power remains for critical facilities like hospitals and water plants, ensuring the grid stays protected.  

The Role Of Sovereign AI And Domestic Silicon 

To support energy independence, the Energy Department requires that both hardware and AI model weights be produced in the US. The AI runs on high-performance computing with specialized accelerators for extreme temperatures. This sovereign AI approach avoids supply chain risks and keeps grid control in domestic hands.  

The system also uses explainable AI protocols, which are methods that show how AI makes decisions for every automated action, such as shutting down a transformer to prevent a wildfire or redirecting power during a heatwave. The system creates a logic trace, a record of the AI’s reasoning. This lets engineers review the AI’s decisions later to ensure its actions align with public safety and long-term goals.  

Integrating the Hydrogen and EV Ecosystems 

As the US works toward its 2050 net-zero goals, the power grid must coordinate with the growing hydrogen sector and the rise of electric vehicles. The AI system acts as an active market operator, managing the two-way flow of energy between the grid and EV batteries, including vehicle-to-grid (V2G).  

When there is surplus wind power at night, AI can automatically lower electricity prices for hydrogen plants and EV charging stations, absorbing excess energy. In the evening, the system draws small amounts from many car batteries to help meet demand until solar returns. This coordination makes the grid a flexible battery, maximizing every green electron.  

Addressing the Human Factor in the Control Room 

This deployment does not mean human grid operators are no longer needed. Instead, their roles are changing as routine tasks are automated, such as frequency and voltage control. The AI enables experts to focus on strategic planning and long-term maintenance. The DOE has started a training program to help current operators become grid architects who can work with the AI to design the next generation of resilient power systems.  

The Life Of A Thinking Continent 

As these digital systems connect across the country, the grid is becoming more intelligent and responsive. It is no longer just wires and transformers. The system now adapts to daily needs. The grid can respond to demand swings like a city turning off its lights or an industrial center starting up at night. In the future, power lines may become the nervous system of smart infrastructure, keeping homes powered and safe.

Source:  Gridmind powering the control room of the future with ai agents

IBM Partners With Eleven Labs to Add Natural Voice to Enterprise AI Systems

ElevenLabs and IBM announced a collaboration to bring ElevenLabs Text-to-Speech and Speech-to-Text (STT) technology to IBM WatsonX Orchestrate. This partnership equips clients with tools for natural language interactions that enhance AI-driven experiences while meeting enterprise security and scalability needs.  

Natural voice is key for customers and employees interacting with AI, but robotic voices, rigid flows, and long waits hurt the experience. ElevenLabs’ advanced TTS enables secure, natural-sounding voice agents in 70 languages that convey human emotion.  

AI agents are becoming central to everyday work, and voice is where AI either earns trust or loses it, said Mati Staniszewski, co‑founder at ElevenLabs. Together with IBM, we are helping organizations replace robotic interactions with AI agents that people actually want to talk to, built with the security and compliance controls that enterprises require.  

This integration enables organizations to use high-quality, multilingual AI voice agents, enhancing experiences across sectors such as government, banking, insurance, healthcare, and utilities in customer service and internal operations.  

IBM Watsonx Orchestrate is the platform for building, deploying, managing, and overseeing AI agents to automate workflows. It connects with existing systems, models, or automation tools, enabling agents to collaborate and providing a solid foundation for enterprise AI clients. Clients get access to ElevenLabs, high-quality speech, and a library of 10,000-plus voices, plus protections including PCI compliance, zero retention mode for HIPAA compliance, and data residency. This ensures consistency, security, and reliability for large deployments and high-volume global interactions.  

We are giving AI agents a voice in the enterprise. Clients are increasingly deploying agentic AI with customer and employee interactions. They want these experiences to be intuitive, responsive, and accessible, said Nick Holder, Vice President, AI Technology Partnerships at IBM. IBM’s open-ecosystem approach gives clients the flexibility to choose the models and tools that best fit their business. Our integration of ElevenLabs with WatsonX Orchestrate shows this. It enables enterprises to deploy AI agents that sound natural, scale globally, and address security, reliability, and governance.  

ElevenLabs and IBM plan to keep working together to help businesses move from text-only agents to voice-first, human-centered AI experiences that can grow with their needs. To discover how your organization can benefit from these advancements, explore IBM WatsonX Orchestrate with ElevenLabs today.  

Statements regarding 11 Labs and IBM’s future direction and aims are subject to change or withdrawal without notice and represent only goals and objectives.  

About ElevenLabs 

Eleven Labs is an AI research and product company changing how we communicate with technology. We launched in January 2023 with the first human-like AI voice model. Today, we serve millions of users and thousands of businesses across three main platforms. ElevenAgents enables businesses to deliver flawless intelligent customer experiences through integrations, testing, monitoring, and the reliability needed to deploy voice and chat agents at scale. ElevenCreative empowers creators and marketers to generate and edit speech, music, image, and video across 70+ languages. ElevenAPI gives developers access to our leading AI audio core models.  

About IBM 

IBM is a leading provider of global hybrid cloud, AI, and consulting expertise. We help clients in more than 175 countries capitalize on insights from their data, streamline business processes, reduce costs, and gain the advantage in their industries. Thousands of governments and corporate entities in critical infrastructure areas such as financial services, telecommunications, and healthcare rely on IBM’s hybrid cloud platform and Red Hat OpenShift to effect their digital transformations quickly, effectively, and securely. IBM’s breakthrough innovations in AI, quantum computing, industry-specific cloud solutions, and consulting deliver open and flexible options to our clients. All of this is backed by IBM’s strong, long-standing devotion to trust, transparency, responsibility, equity, and service. Visit www.ibm.com for more information. 

Source: Enterprise AI Finds its Voice: ElevenLabs and IBM Bring Premium Voice Capabilities to Agentic AI 

Apple Opens US Orders for AirPods Max 2 With On-Device AI Features

Apple today announced AirPods Max 2. These headphones deliver clearer calls, richer sound, and industry-leading noise cancellation, providing users with more immersive listening and recording experiences. With the H2 chip, AirPods Max 2 introduces adaptive audio for seamless transitions between environments, live translation for real-time communication, and a studio-quality recording feature, making them especially valuable for podcasters, musicians, and content creators.   

Be among the first to experience AirPods Max 2 by ordering yours starting March 25 at midnight in starlight, orange, purple, or blue. Act now to ensure early access next month.  

With the H2 chip, AirPods Max are upgraded. ANC is 1.5 times more effective, said Eric Treski, Apple’s director of audio product marketing. Sound quality is clean and detailed, and personalized spatial audio enhances the experience.  

Active Noise Cancellation Gets Even Better 

With H2 and new computational audio, AirPods Max 2 cut external noise by up to 1.5 times more than before. This lets users enjoy clearer music, conversations, and focus, even in busy airports or on trains.  

A new digital signal processing algorithm, organized for H2 and the AirPods Max microphone array, delivers more natural transparency, helping users remain aware of their surroundings.  

Highest Fidelity Listening Experience 

AirPods Max 2’s high-dynamic-range amplifier delivers cleaner, more detailed sound, letting users hear music, movies, and games with pinpoint instrument location and balanced bass, mids, and highs.   

AirPods Max 2 supports lossless audio and can be used professionally, with personalized spatial audio available via USB‑C.  

Reduced wireless audio latency ensures games feel responsive and immersive, giving users a competitive edge and smoother experience on iOS, macOS, and iPadOS devices.  

Intelligent features come to AirPods Max. 

The H2 chip brings new features that make AirPods Max 2 smarter and more helpful: users get instant audio optimization, clearer calls, and easier communication across languages for a seamless daily experience.  

  • Adaptive audio automatically adjusts ANC and transparency levels based on the environment to optimize the audio experience.  
  • Conversation awareness lowers content volume and reduces background noise when the user begins speaking to someone.  
  • Live translation is powered by Apple Intelligence. It enables users to communicate across languages in person.  
  • Voice isolation uses advanced computational audio with H2. It focuses on users’ voices during calls and blocks ambient noise.  
  • With the camera remote, users can take photos or start and stop video recording from a distance by pressing the Digital Crown on their Apple Watch. This feature works with the Camera app or compatible third-party apps on iPhone. It also enables creators to record content with higher-quality audio and a more natural vocal texture.  
  • Loud sound detection helps prevent exposure to loud environmental noise while preserving the audio’s sound signature.  
  • Personalized volume automatically fine-tunes the listening experience over time based on user preferences.  
  • Siri interactions enable users to respond to Siri announcements by nodding their head for yes or gently shaking their head for no.  

AirPods, Max 2, And The Environment 

Apple 2030 is the company’s plan to achieve carbon neutrality across its entire footprint by the end of the decade, reducing product emissions from materials, electricity, and transportation. AirPods Max 2 use 100% recycled rare-earth elements in all magnets, 100% recycled polyester in the ear cushions, and 100% recycled gold plating and tin solder on all Apple-designed printed circuit boards. The paper packaging is entirely fiber-based and can be easily recycled.

Source: Apple introduces AirPods Max 2 

Intel Launches Core Ultra vPro Chips Built on 18A for Enterprise AI PCs

News Highlights 

  • The Intel Core Ultra Series 3 for business PCs, built on Intel 18A, delivers advanced performance and efficiency for commercial computers, powering everything from lightweight business laptops to powerful workstations.  
  • The Intel vPro platform provides advanced security, AI-powered management, and easy fleet service activation used by over 1,300 commercial customers worldwide.  
  • The new Intel Arc Pro B-series GPUs provide cost-effective, high-performance graphics solutions for content creation, engineering, and AI applications.  
  • Intel Xeon 600 workstation processors are now available, offering scalable performance for professional and technical computing.  

New York, March 25, 2026 Intel has introduced its most advanced commercial client portfolio designed for all types of professionals and supporting over 125 different designs: the new Intel Core Ultra Series 3 vPro offers users efficient performance, strong security, and easy device management for IT teams. Intel also announced its high-end Intel Arc Pro B70 and B65 graphics cards, as well as the retail launch of Intel Xeon 600 workstation processors. These products are built to deliver high computing power and scalability for professionals, solving problems across any industry.  

The Intel Core Ultra Series 3, the first commercial PC platform on 18A, offers next-gen performance and built-in AI acceleration with Intel vPro. Organizations gain easy deployment, proactive, effective security, and smarter fleet operations.  

From laptops to workstations, this is Intel’s most expansive commercial portfolio, enabling IT and business leaders with the performance, efficiency, security, manageability, and AI capabilities needed for the next era of work. David Feng, vice president of Client Computing Group and GM of PC segments.  

Intel Core Ultra Series 3 Processors Built for Business 

The Intel Core Ultra Series 3 offers strong performance, better power efficiency for longer battery life, new graphics, and AI acceleration for business computers.  

Compared to 4-year-old systems, the average PC refresh business users can expect.  

  • Over 30 percent faster single- and multi-thread performance.  
  • Up to 30 percent higher productivity compared to the average 4‑year‑old business Pc’s.  
  • Up to 80 percent better graphics performance versus average 4-year-old business Pc’s.  
  • Up to 4x AI performance versus average 4-year-old business PCs.  

Together, these improvements help users work more efficiently, collaborate more effectively, and use AI-powered workflows across the company.  

Intel vPro: The Foundation of Modern Managed PCs 

Intel vPro continues to be a leader in business security and fleet management, offering built-in security, easy management, and trustworthy performance.  

New Capabilities Include 

  • The Intel vPro certification program boosts application and accessory performance by working closely with software vendors and partners. This reduces CPU usage, improves power efficiency, and reduces background activity. Early results show up to 59% less CPU use, 56% better power efficiency, and 74% less background activity.  
  • Intel vPro Intelligence with Device IQ uses AI-powered analytics to find, diagnose, and fix issues before they escalate. This helps reduce downtime and the need for IT support. Integration with digital experience (DEX) platforms is planned for the second half of 2026.  
  • Intel vPro Fleet Services makes out-of-band management and disaster recovery easier with the first fully managed turnkey SaaS-based activation. It is also the first silicon partner to work with Microsoft Intune, so there is no need for extra infrastructure. IT teams can quickly activate Intel vPro management from the Intune admin center, speeding up deployment and simplifying operations. Over 1,300 commercial customers worldwide started using the service last quarter.  
  • Intel vPro Security Enhancements now include Intel Total Storage Encryption for Microsoft BitLocker and AI-based Threat Detection with Intel Threat Detection Technology (DTECT). This is the only silicon that can detect the most advanced malware threats in real time.  

Intel Arc Pro Graphics Built for Creators and AI Developers 

The Intel Arc Pro B70 and B65 discrete GPUs provide high-performance, value-driven professional graphics for advanced content creation, engineering tasks, and demanding AI workloads.  

Built on the Xe2 architecture, these GPUs feature up to 32 Xe cores and 32 GB of VRAM and are optimized for multi-user AI workloads.  

  • They are optimized to handle AI workloads involving multiple users and agents.  
  • They offer a strong balance of price and performance for AI inference, whether employed in workstations or edge devices.  

Performance Benefits Include 

  • Up to 2.2x larger context windows with the Intel Arch Pro B70 compared to similar competitor graphics cards.  
  • Up to 6.2x faster responses in multi-agent/multi-user workloads with the Intel Arc Pro B70 compared to similar competitor graphics cards.  
  • Up to 2x tokens per dollar performance with the Intel Arc Pro B70 versus the competition.  

Intel Arc Pro has evolved from powering creative workflows to enabling AI builders at scale. With the Intel Arc Pro B70 and B65, we are providing high-performance workstation graphics products and AI inference at exceptional value. Anil Nanduri, Vice President, AI Products and GTM, Intel Data Center Group.  

Availability 

These commercial pieces powered by Intel Core Ultra Series 3 and Intel vPro will be available starting March 31, 2026.  

The Intel Arc Pro B70 discrete GPU will be available starting March 25, 2026, both as an Intel-branded card model and through Intel AIB partner cards from ARKN, ASRock, Gunnir, Maxsun, Sun, and Sparkle. Intel-branded Arc Pro B70 discrete GPUs will have a suggested starting price of $949. Pricing for Intel Arc Pro B70 discrete GPUs from AIB partners will vary based on final model configurations.  

Intel Arc Pro B65 discrete GPUs will be available starting in mid-April 2026 through Intel’s AIB partner network with a wide range of form factors. Intel Arc Pro B65 pricing will vary based on final model configurations.  

Intel Xeon 600 processors for client workstations will be available starting March 25, 2026. Suggested component pricing ranges from $499 to $7,699, and end-user system pricing depends on the final component configuration.

Source:  Intel Core Ultra Series 3 with Intel vPro Powers Next Generation of Commercial PCs Built on Intel 18A

CrowdStrike and AWS Name, AI Startup Winner in 2026 Security Accelerator

CrowdStrike announced at RSE 2026 that Jazz won the third annual Cybersecurity Startup Accelerator held in partnership with Amazon Web Services (AWS) and NVIDIA. Through the NVIDIA Inception program, JAIS was recognized for its novel AI-powered data loss prevention (DLP) approach and its promise to solve key security challenges for today’s enterprises.  

Nearly 1,000 startups from around the world applied to the 2026 accelerator. This interest shows a growing demand for AI-powered cloud security. Thirty-five startups joined the eight-week, equity-free program. They collaborated with experts from CrowdStrike, AWS, NVIDIA, and others to transform ideas into enterprise-ready security platforms and launch their businesses.  

On March 24, 2026, at the RSAC-2026 conference in San Francisco, Jazz was named the winner of the Cybersecurity Startup Accelerator. The choice was made from six finalists. The judges included George Kurtz, CEO and founder of CrowdStrike; CJ Moses, Chief Information and Security Officer (CISO) at Amazon; Richardson, senior director of Agentic AI and cybersecurity engineering at NVIDIA; and special guest judge Robert Herjavec, entrepreneur and Shark Tank judge.  

Above Security was named the runner-up for its use of AI agents, software that uses artificial intelligence to perform tasks autonomously to manage insider risk and security threats within an organization, and to reduce traditional alerts with detailed investigative reports.  

The accelerator highlights a trend seen at both the RSA Conference, a major cybersecurity event, and NVIDIA GTC 2026, NVIDIA’s AI and graphics conference. Hardware companies, enterprise service providers, and new founders are building on a shared security platform. This platform offers technology and infrastructure designed for collaborative deployment and protection.  

We are incredibly grateful to be the winner of the CrowdStrike, AWS, and NVIDIA cybersecurity startup accelerator. Said Ido Livneh, co-founder and CEO at JAIS: “DLP has been broken for decades by rule writing, alert floods, and no real answers. So, we rebuilt it from the ground up using first principles thinking as an AI-native system to truly understand how data moves through a business.” The advice and assistance we received through the accelerator were essential to sharpening the product and validating it against real-world enterprise demands as we scale.  

The path to becoming a defining cybersecurity company looks nothing like it did just three years ago, said Daniel Barnard, chief business officer at CrowdStrike. Today’s promising founders are building on leading platforms from the start. Together with AWS and N media, this accelerator supports early innovation. JAWS stood out for its context-aware, AI-native DLP approach, leveraging Melody and Context Vault to provide clear, actionable answers at scale and exemplify leadership in the new cybersecurity era.  

Security innovation only matters if it performs within real cloud environments, said CJ Moses, CISO of Integrated Security at Amazon. This accelerator connects founders directly with the infrastructure, operational expectations, and scale enterprises demand. What sets Jazz apart is its ability to investigate signals with full business context and surface meaningful incidents with minimal analyst overhead, meeting the standards modern security teams expect.  

The shift to agentic AI is transforming cybersecurity, requiring autonomous systems that can perceive, reason, and act against increasingly sophisticated threats, said Bartley Richardson, senior director of agentic AI and cybersecurity engineering at NVIDIA. The accelerator equips the next generation of security platforms with high-performance compute and foundational frameworks needed for secure, scalable AI. Judge distinguished itself by its agentic investigator, Melody. It can analyze multidimensional context across data, systems, people, and businesses. This helps drive precise, in-context prevention decisions and distills millions of events into a small set of usable insights.  

For more information, visit the program’s website. 

Source: CrowdStrike and AWS Announce Jazz as the Winner of the 2026 Cybersecurity Startup Accelerator, Supported by NVIDIA Inception 

Microsoft Defender Adds Identity Risk Scoring for AI-Driven Security Decisions

Next week, the RSAC conference celebrates its 35th anniversary as the security community’s pivotal event for meeting new challenges and discovering opportunities to make the world safer. As we reach this milestone, Agentic AI is quickly changing industries. 

Many customers are becoming frontier firms focusing on intelligence and trust and using agents to help people reach their goals and rethink how they do business. Our recent research shows that 80% of Fortune 500 companies are already using agents.  

At the same time, as these innovations expand, we are also seeing a rise in AI-powered attacks in which agents can act as double agents. CIOs, CISOs, and other security leaders now face important questions. How can they monitor, manage, and secure agents? How do they protect their core systems in this new era? And how can agentic AI help defend their organizations against both old and new threats?  

To address these challenges, the solution begins with trust, and security is the foundation of that trust. In this new era of agentic AI, security needs to be built into every part of the AI system. It should work in the background and on its own. Like the AI, it protects. This is our vision: making security the core of the AI stack.  

At RSAC 2026, we are bringing this vision to life with new tools that help organizations secure agents, protect core systems, and defend themselves with support from agents and experts. Microsoft Security processes over 100 trillion signals daily, protecting 1.6 million customers, 1 billion identities, and 24 billion Co-Pilot interactions. Keep reading to see how we can help you secure agentic AI.  

Secure Agents 

Agent 365, available May 1, centralizes agent management for IT, security, and business teams using existing infrastructure. It includes the new Defender Era and Purview features for secure access, data protection, and threat defense.  

Agent 365 is part of Microsoft 365 E7, the Frontier Suite, which also includes Copilot, Intra-Suite, and E5, offering advanced air security features for full protection.  

Secure Your Foundations 

In addition to securing agents, AI must also be protected at every level. Securing agent AI requires safeguarding its systems and the people who build and use it. At RSAC 2026, we are launching new features to help you spot organizational risks, secure identities and adaptive access controls, protect sensitive data in AI workflows, and defend against evolving threats.  

Gain Visibility Into Risks Across Your Enterprise 

As AI adoption accelerates, so does the need for comprehensive continuous visibility into AI risks across your environment. As more organizations adopt AI, it becomes even more important to have clear, ongoing visibility into AI risks across your environment—from agents to AI apps and services. We are meeting this need with new tools that show you where AI is being used, how it is used, and where your risks might be increasing. These new features are now generally available.  

  • Entra Internet Access Shadow AI Detection identifies unknown and unmanaged AI apps through network analysis. Available March 31.  
  • Enhanced Intune App Inventory provides rich visibility into the apps installed on your devices, including AI-enabled apps to support targeted remediation of high-risk software, generally available in May.  

Secure Identities With Continuous Adaptive Access 

Identity is the foundation of modern security, often the primary target in any system, and the first line of defense against threats. With Microsoft Entra, you can secure access and strengthen identity protection using new features that help you improve your identity setup, manage tenants better, update authentication methods, and make smarter access decisions.  

  • Entra backup and recovery strengthens resilience by automatically backing up Entra directory objects, enabling rapid recovery in the event of accidental data deletion or unauthorized changes. Now available in preview.  
  • Entra-Tenant Governance discovers and manages shadow tenants with consistent policies in preview.  
  • Entra passkey capabilities now include synced passkeys and other passkey profiles, enabling maximum flexibility for end users and making it easy to move between devices, while organizations seeking maximum control can still use device-bound passkeys. In addition, intra-passkeys are now natively integrated into Windows Hello, making phishing-resistant passkey authentication even more seamless across Windows services. Synced passkeys and passkey profiles are generally available. Passkey integration in Windows Hello is in preview.  
  • Entra External Multi-Factor Authentication allows organizations to connect external MFA providers directly to Microsoft Entra, enabling them to leverage pre-existing MFA investments or use highly specialized MFA methods that are now generally available.  
  • Entra Adaptive Risk Remediation helps users securely regain access without help desk friction by automatically self-remediating across authentication methods, adapting to where they are in their modern authentication journey. It will be generally available in April.  
  • Unified Identity Security offers complete protection across your identity systems, control center, and threat identification and response, all designed for quick action and instant decisions. The new identity security dashboard in Microsoft Defender displays the most important data for both human and non-human identities, helping you respond faster. The new Identity Risk Score combines risk signals from different accounts to provide a comprehensive view of user risk, supporting real-time access decisions and security investigations. Now available in preview.  

Secure Sensitive Data Across AI Workflows 

As AI becomes part of daily work, sensitive data now moves quickly through prompts, responses, and grounding flows, sometimes outpacing existing policies. Security teams need to see how AI uses data and be able to prevent oversharing or leaks. 

Microsoft now brings data security into the AI control pane, giving organizations a clear view of risks, real-time enforcement, and the confidence to use AI responsibly across the business. New Microsoft Purview features include.  

  • Expanded Purview Data Loss Prevention for Microsoft 365 Copilot helps block sensitive information such as PII, credit card numbers, and custom data types from being processed or used for web scraping. Generally available March 31.  
  • Purview embedded in the Copilot control system provides a unified view of AI-related data risk directly in the Microsoft 365 admin center. Generally available in April.  
  • Purview customizable data security reports enable tailored reporting and drilldowns to prioritized data risks. available in preview on March 31.  

Defend Against Threats Across Endpoints, Cloud, and AI Services 

Security teams need round-the-clock protection that can stop threats early and automatically contain them. Microsoft is expanding predictive shielding to limit risks and reduce exposure, improving container security, and adding network-level protection against harmful AI prompts.  

  • Entra, internet access, prompt injection protection helps block malicious AI prompts across apps and agents, enforcing universal network-level policies. Generally available March 31.  
  • Enhanced Defender for cloud container security includes binary drift and anti-malware protection to close gaps that attackers exploit in container enforcements, now available in preview.  
  • Defender for Cloud Posture Management adds broader coverage and supports Amazon Web Services and Google Cloud Platform, delivering security recommendations and compliance insights for newly discovered resources, now available in preview in April.  
  • Defender Predictive Shielding automatically adjusts identity and access policies during active attacks. Reducing Exposure and Limiting Impact is now available in preview.  

Defend With Agents And Experts 

To protect organizations in this new era, a defense approach designed for agents is needed. This includes deploying a defense platform and security agents embedded in daily workflows, all supported by expert knowledge and robust security services when required.  

Agents Are Built Into the Security Workflow 

Security teams work best when they get help, right where and when they need it. As alerts arise and investigations span identities, data, devices, and cloud networks, AI-powered tools should directly support defenders with Security Copilot. 

Now part of Microsoft 365 E5 and E7, we are giving defenders agents that are built into daily security and IT tasks. These agents help speed up responses and reduce manual work, so teams can focus on what matters most.  

New Agents Available Now Include 

  • Security Analyst Agent in Microsoft Defender helps accelerate threat investigations by providing contextual analysis and guided workflows. Available in preview on March 26th.  
  • The security alert triage agent in Microsoft Defender has the capabilities of the phishing triage agent and extends to cloud and identity, autonomously analyzing, classifying, prioritizing, and resolving repetitive low-value alerts at scale. Available in preview in April.   
  • Conditional Access Optimization Agent in Microsoft Intra Enhancements: Add context-aware recommendations, deeper analysis, and staged rollout to strengthen identity security. Agent is generally available, and enhancements are now available in preview.  
  • Data Security Posture Agent enhancement in Microsoft Purview includes a credential-scanning capability that can proactively detect credential exposure in your data. Now available in preview.  
  • Data Security, Tri-Age Agent Enhancements in Microsoft Purview include an advanced AI reasoning layer and improved interpretation of custom sensitive information types (SITs) to enhance agent outputs during alert triage. The agent is generally available, and enhancements are available in preview from March 31.  
  • Over 15 part-time built agents extend Security Copilot with additional capabilities, all available in the Security Store.  

Scale With an Agentic Defense Platform 

To help defenders and agents work together more smoothly and intelligently, Microsoft is expanding Sentinel as a defense platform. This update delivers context, automates workflows from start to finish, and standardizes access, governance, and deployment across security tools.  

  • Sentinel Data Federation, powered by Microsoft Fabric, investigates external security data in Databricks, Microsoft Fabric, Azure, and Azure Data Lake Storage while preserving governance. Now available in preview.  
  • The Sentinel Playbook Generator with natural-language orchestration helps accelerate investigations and automate sophisticated workflows. Now available in preview.  
  • Sentinel, Granule, Delegated Administrator privileges, and unified role-centric access control enable secure and scalable management for partners and enterprise customers, with cross-tenant collaboration now available in preview.  
  • Security store embedded in Purview and Entra makes it easier to discover and deploy agents directly within existing security experiences. Generally available March 31.  
  • Sentinel custom graphs powered by Microsoft Fabric enable views unique to your organization of relationships across your environment. Now available in preview.  
  • The Sentinel Model Context Protocol (MCP) Entity Analyzer helps automate faster with natural language and harnesses the flexibility of code to accelerate responses generally available in April.  

Strengthen with Experts 

Even the most experienced security teams sometimes need extra support, especially during complex attacks or investigations. The Microsoft Defender Experts suite offers expert-led services, including technical advice, managed external detection and response, MXDR, and full incident response. These services help you defend against advanced threats, build sustained resilience, and modernize your security operations with confidence.  

Apply Zero Trust for AI 

Zero Trust centers on three ideas: verify consistently, use the least privilege, and assume breaches as AI permeates the environment from models to agents. These principles are critical. At RSAC 2026, we are broadening our Zero Trust architecture to cover the full AI lifecycle. Our updated reference architecture workshop, assessment tools, and new articles make this practical for you. 

Source: Secure agentic AI end-to-end 

Google Cloud Launches AI Service to Detect Threats From Dark Web Activity

Our News Today From the RSA Conference 

  • We have completed the acquisition of Wiz, empowering cybersecurity teams across multiple cloud service providers in today’s AI-driven world.  
  • New research from Mandiant’s M-Trends 2026 and our AI Risk and Resilience Report helps organizations stay up to date on threats.  
  • We are adding new agents to the agent SOC, helping defenders automate tasks, respond faster, and focus on critical security events.  
  • Check out our latest security updates in Chrome Enterprise, Security Command Center, Network Management, and more.  

AI-driven defense is transforming cybersecurity in ways defenders have long wanted. Google Security is bringing its strongest tools yet to the RSA Conference, building on the Agentic Security Operations Center by leveraging Gemini models to automate reasoning. The agentic SOC equips defenders to detect, investigate, and respond to threats more effectively, giving them a significant operational advantage.  

Today, we are providing updates across our products, introducing new developments with Wiz. The release of M-Trends 2026, which contains insights from Mandiant’s investigations and important changes in how we use threat intelligence. Keep reading to see how Google Security can help you stay ahead.  

Welcoming Wiz to Google Cloud 

Google has officially acquired Waze. Together, we will offer an AI-ready cybersecurity platform for all environments.  

We believe that making multi-cloud security simpler lets you innovate with confidence, no matter where your data and apps are. We are excited to show you how Waze helps organizations quickly and safely deploy AI with their AI Application Protection Platform and AI App, and how security teams can work faster with their red, blue, and green security agents. Learn more about our common mission from Google Cloud CEO Thomas Kurian.  

M Trends 2026 – Useful Insights From 500k Plus Hours of Incident and Investigations 

M-Trends 2026 helps organizations understand the evolving threat landscape and defense strategies, highlighting the need to observe faster initial access and longer-term intrusions.  

Adversaries are no longer just stealing data. Cyber criminals are increasingly operating like highly efficient businesses, creating partnerships that have collapsed the window for defenders to intervene from hours to just 22 seconds. They want to completely dismantle an organization’s ability to restore operations while increasing their leverage in extortion. Download today for useful insights.  

We recently released a new Mandiant report on AI risk and resilience. This report, based on 2025 Mandiant Consulting and Google Threat Intelligence Group data, details how adversaries have quickly moved from testing AI to deploying adaptive tools and autonomous agents that leverage AI for real-time code rewrite. To address risks like shadow AI and incomplete asset visibility, organizations should move beyond passive oversight by adopting ongoing red teaming and stress testing while leveraging the speed and analysis of AI-powered defense.  

Users of Google Security Operations can add agents, such as our triage and investigation agent, directly into their procedures to speed up response times. The triage and investigation agent automatically investigates alerts, collects evidence, and gives verdicts with clear explanations.  

This information helps security analysts automate decision-making, close alerts, and manage fixes, so they can focus on the more important threats rather than false positives. Building workflows that use this agent will make it even easier for security teams to coordinate their response.  

Few would argue that the progress made over the last 12 to 18 months in putting AI to work to improve security operations is remarkable. New research from Omdia shows that 89% of CISOs are pushing to accelerate the adoption of agentic security, said David Gruber, principal analyst, cybersecurity, Omdia. 

Not only does this commitment reflect the immediacy of combating an AI-enabled adversary, but our data also show that over half of cybersecurity practitioners believe that authentic AI offers a greater advantage to cybersecurity defenders than the adversary. With the prospect of significant improvements in security outcomes, Google Cloud is well-placed to help organizations transform their SOCs with this powerful new technology. In Google Security Operations, customers can now create their own enterprise-ready security agents managed by the remote MCP server, arriving in April. This eliminates the need to host your own MCP server client, simplifying management. build.   

Decision To Use Dark Web Intelligence 

Most threat intelligence teams today spend their days sorting through too many low-quality alerts. The main problem is not a lack of information but a lack of relevant information. To clarify intelligence and find hidden threats, we’ve added agentic features to Google Threat Intelligence. AI agents using Gemini models handle analysis, letting analysts focus on what matters.  

To help teams shift from manual triage to agentic defense, we are adding Dark Web Intelligence to Google Threat Intelligence. Our GTIG analysts, with deep experience in the dark web, provide important context for Gemini’s capabilities. The new feature uses the latest Gemini models to automatically create a detailed profile for your organization.  

Our internal tests show that it can review millions of daily external events with 98% accuracy, highlighting only the threats that matter to your mission. By providing clear answers that explain the why and how of a threat, we help defenders save time and stay ahead in a world of increasingly automated threats.  

Customers can now turn large amounts of dark web data into clear, relevant insights delivered quickly with AI. This helps your team think and act faster than opponents using agents.  

In previous roles, I have leveraged several dark web tools and found that they yielded false-positive rates of over 90%. The new dark web intelligence filters out this noise and connects the dots that no human analyst can see in time. It’s the difference between reacting to a fire and putting it out before a match is struck, said Michael Kosak, director of Threat Intelligence at LastBy. Moving from simple keyword matching to intent-based analysis, dark web intelligence can better understand the context of an adversary’s actions. For example, it can spot when a subsidiary’s access is compromised even if the attacker does not name the victim.  

Protecting your AI innovation 

You need agentic defense to protect your organizations at machine speed, and you also need to protect your AI innovation. As organizations move from testing AI to using it at scale, a big confidence gap has appeared. 72% of organizations are not confident in their ability to run a secure AI strategy, according to a recent survey by Cloud Security Alliance (CSA) and Google.  

Google Cloud can help close this gap by supplying a complete approach to securing AI innovation. We protect the whole life cycle from building to running and cover everything from infrastructure and data to models and agents.  

To help with these problems, we are offering customers new key features.  

  • AI protection in the Security Command Center now integrates with the Vertex AI Agent Engine to detect agent threats, such as unauthorized entry and data exfiltration attempts.  
  • New Armor now integrates with Google NCP servers, expanding its coverage to help mitigate agentic risks, including direct and indirect prompt injections, sensitive data leakage, and tool poisoning.  
  • Sensitive data protection adds AI-powered context classification, medical finance, and passport object detection.  
  • Security Command Center: external export management (available soon in preview) will provide SCC users with a validated outside-in view of your Google Cloud attack surface, identifying exploitable vulnerabilities and showing the native network path enabling the exposure.  

Switching gears to network safeguards, here’s what’s new in network security. 

Google Cloud Network Security has new capabilities for protecting apps and enforcing policies across clouds.  

  • Network security integration in band mode secures app workloads with third-party appliances. No routing changes required.  
  • Cloud NGFW Regional Firewall Policies Preview Protect Workloads Via Internal Application and Proxy Network Load Balancers  
  • Cloud Armor adds hierarchical policies and organization-wide address groups for centralized management and stronger defenses, sets WAF rule limits, configures policies at multiple levels, and manages IP lists across policies.  

What’s new in Chrome Enterprise Premium? 

Chrome Enterprise Premium still protects organizations from data loss. At the RSI conference, we are showcasing new improvements and integrations with our partner Citrix.  

  • Enterprises already benefit from Chrome Enterprise protections against unauthorized use of AI tools in the browser. Now Citrix and Chrome Enterprise together offer even more defense for shared customers, including key‑logging protection and ongoing device checks.  
  • Clipboard protection now extends to Citrix Virtual Apps and web-based apps. Chrome Enterprise’s new browser cache encryption provides added security for non-corporate devices.  

Join Google Security At RSAC 2026 

Our experts are ready to connect and work with you. Visit us to see our technology in action at Moscone’s North Hall booth N6062 or at our space at the Marriott Marquis. You can also engage with the future of cybersecurity in over 19 sessions we are hosting.  

Find out how you can make Google part of your security team. If you cannot join in person, you can live-stream RSAC content or watch it later on demand.

Source: RSAC ’26: Supercharging agentic AI defense with frontline threat intelligence 

Meta and Arm Partner on Custom Data Center Chips for AI Workloads

Takeaways 

  • We are working with ARM to create a new CPU for data centers and large-scale AI projects.  
  • With ARM, we will develop multiple generations of CPUs, expanding our silicon portfolio and enabling more AI experiences.  

We are partnering with ARM to co-develop a CPU specifically architected for the computational intensity of AI training and inference in data centers. As AI workloads accelerate, data centers demand CPUs with increased processing power and efficiency to support large, distributed AI deployments. This collaboration will yield processors optimized for high-throughput, energy-efficient operation critical to Meta’s scalable AI infrastructure.  

Delivering AI experiences at a global scale demands a robust, adaptable portfolio of custom silicon solutions purpose-built to accelerate AI workloads and optimize performance across Meta’s platforms. We worked alongside ARM to develop the AGICPU to deploy an efficient compute platform that significantly improves our data center performance density and supports a multi-generation roadmap for our evolving AI CPS systems.  

Santosh Janardhan, Head of Infrastructure, Meta.  

Arm has introduced the AGI CPU, a production-ready chip built on Arm’s Neoverse platform to support the next generation of AI infrastructure.  

For the first time in over 35 years, ARM is releasing its own silicon chips  actual physical chips. This expands ARM’s Neoverse platform beyond just IP (intellectual property) and technology designs for compute subsystems. It completes the building-block process, giving customers more ways to use ARM technology. Now they can build custom silicon, use platform-level solutions, or choose ARM-designed processors. This move responds to the fast-changing AI infrastructure and the growing demand for ready-to-use ARM platforms that can be quickly deployed at scale.  

The Rise Of The Agentic AI Infrastructure 

AI systems now run nonstop globally. Previously, human interaction limited computing speed. Agentic AI removes this bottleneck by enabling software agents to orchestrate tasks, leverage multiple models, and make decisions instantly.  

As AI workloads become more multifaceted, CPUs are essential for maintaining reliable operation across distributed AI systems in data centers. They direct multiple tasks while coordinating accelerators, memory, storage, and software agents at scale.  

These changes put new demands on CPUs, so processors must evolve.  

ARM Neoverse powers the leading hyperscale and AI platforms, including AWS, Graviton, Google, Axon, Microsoft, Azure, Copart, and NVIDIA, Vera. As AI infrastructure grows, partners ask ARM for more. The ARM AGI CPU meets this need.  

This Is ARM AGI CPU Built for Rack Scale Agentic Efficiency 

Agentic AI workloads need steady, large-scale performance. The ARM AGI CPU is built to deliver strong performance for each task, even under heavy workloads and across thousands of cores, while staying within the power and cooling limits of today’s data centers.  

Every aspect of the ARM AGI CPU from clock rates and memory bandwidth to input/output (I/O) paths is designed for highly parallel agentic AI workloads. The architecture ensures high performance in dense computing environments typical of modern data centers.  

The ARMS reference server uses a 1-OU unit about 1.7 inches tall, a 2‑node setup with 2 chips and dedicated memory and I/O connections, and 272 cores per blade, a server component. These blades fill a standard air‑cooled 36 kW rack with 30 blades providing 8,160 cores in total. Arm also worked with Supermicro on a liquid‑cooled 200 kW design that can host 36 Arm AGI CPUs, totaling more than 45,000 cores. With this setup, the Arm AGI CPU can deliver over twice the performance per rack compared to the latest x86 systems  thanks to the strengths of the Arm architecture and careful alignment of system resources with computing needs.  

  • The ARM EGR CPUs’ high memory bandwidth allows more effective threads per rack, while x86 CPUs slow down as cores compete under heavy workloads.  
  • The high-performance, efficient single-threaded ARM Neoverse V3 CPU cores (processing units that handle one instruction thread at a time) outperform older architectures. So, each ARM thread gets more done inherently more usable threads, and getting more work per thread leads to huge performance gains per rack.  

The first product of ARM’s AGI CPU is its first data center CPU for the AI era. It offers faster performance per rack and is more efficient than older CPUs. Meta is the lead partner. The ARM AGI CPU is ARM’s first data center CPU for the AI era, offering faster performance per rack and more efficiently than previous CPUs. Meta is the lead partner and core developer focused on improving app infrastructure and integrating with our MTIA silicon. The ARM AGI CPU will be available through ARM, and we will open our board and rack designs under the Open Compute project this year, expanding to deliver production silicon CPUs optimized for large-scale agentic AI deployments. Together, we are combining ARM’s foundation of high-performance power-efficient computing with Meta’s infrastructure expertise to support the next generation of AI systems. Rene Haas, CEO, Arm.  

Our ARM CPUs will enhance our silicon portfolio, helping us build a robust, flexible hardware stack to bring AI to billions worldwide. 

Source: Meta Partners With Arm to Develop New Class of Data Center Silicon 

OpenAI Retires Legacy ChatGPT Modes to Standardize New Reasoning Systems

OpenAI has officially retired several older ChatGPT models, including the popular GPT-4o. Starting February 13, 2026, this change helps standardize the platform on the new advanced GPT-5 series.  

This move encourages users to shift to newer models, especially GPT-5.2. These updated models offer better personalization, improved creative capabilities, and fewer unnecessary rejections or overly agreeable responses, aligning with OpenAI’s goal to offer a consistent experience.  

Key Aspects of the Model Retirement 

  • As of February 13, 2026, ChatGPT no longer supports GPT-4O, GPT-4.1, GPT-4.1 mini, OpenAI-04 mini, and the previously announced GPT-5 (instant/slash thinking).  
  • Transition Focus, OpenAI reported that 99.9% of users had already shifted to newer versions, with GPT-5.2 accounting for the vast majority of daily active usage.  
  • API Continuity: This retirement applies solely to the ChatGPT consumer interface. API access to legacy models stays available for now, with developers receiving advance notice before future API sunsets.  
  • Enterprise Access: Business, Enterprise, and Edu customers can continue using legacy models through custom GPTs until April 3, 2026. After this date, these customers will no longer have access to legacy ChatGPT models.  

New reasoning and customization standards: the move to GPT 5.1 and GPT 5.2 combines the best parts of older models, such as GPT 4.0’s warmer personality, into newer systems. These systems are also more flexible. New features include  

  • Selectable base styles. Users can select base tones: e.g., Friendlier, and adjust Warmth and Enthusiasm levels directly in Settings.  
  • Improved reasoning: the new models prioritize advanced analytical skills, coding, and multi-step reasoning over simpler one-shot chatbot functions.  
  • OpenAI is developing an 18-plus model for Child GPT. This model will offer greater freedom and reduced safeguards for adults.  

This consolidation lets OpenAI accelerate innovation in personalization and creative performance. GPT-4 is rolling out now, bringing the latest advances in reasoning and agentic workflows to ChatGPT.  

Alongside the retirement of GPT-5 Instant and Thinking, GPT-4.1, GPT-4.1 Mini, and OpenAI 404 Mini will also be unavailable on ChatGPT after February 13, 2026. There is no API change at this time.  

In particular, while this announcement applies to several older models, GPT-4O deserves special context.  

After we first deprecated it and later restored access during the GPT-5 release, we learned more about how people use GPT-4 in their day-to-day. We brought back GPT-4o after hearing clear feedback from some Plus and Pro users. They told us they needed more time to transition key use cases, such as creative ideation. They also prefer GPT-4’s dialogue style and warmth.  

That feedback helped shape GPT 5.1 and GPT 5.2, leading to better personality, stronger support for creative ideas, and more ways to customize ChatGPT’s responses. You can now choose base styles and tones, like friendly, and adjust settings for warmth and enthusiasm. Our goal is to give people more control over how ChatGPT feels to use, not just what it can do.  

We are announcing the retirement of GPT-40 now because improved models are available, and most users have already switched to GPT-5.2, with only 0.1 percent still using GPT-40 daily.  

More generally, we are working to improve ChatGPT in areas users have identified as needing attention. This includes improving character and creativity, and reducing unnecessary refusals caused by overly cautious responses. Updates are coming soon. We are also developing a version of ChatGPT for adults over 18, focused on giving adults more choice and freedom with suitable safeguards to help us. To help with this, we have introduced age prediction for users under 18 in most markets.  

We know changes like this take time to get used to. So, we are providing advance notice. Model access will change on February 13, 2026, and limited enterprise access will end on April 3, 2026. We understand that losing access to GPT-40 may be frustrating, and this decision was made with care. Retiring models helps us focus on improving the ones most people use now.

Source:   Retiring GPT‑4o, GPT‑4.1, GPT‑4.1 mini, and OpenAI o4-mini in ChatGPT